Abstract: Objective　To investigate the spatial?temporal characteristics of reported schistosomiasis cases in China from 2004 to 2017, so as to provide insights into the development of different schistosomiasis control strategies at various stages. Methods　The monthly data of reported schistosomiasis cases at a provincial level of China from 2004 to 2017 were collected from the Public Health Science Data Center, and the spatial?temporal distribution of reported schistosomiasis cases was preliminarily identified using a descriptive statistical method. According to the goals at different stages proposed by the National Mid? and Long?term Program for Schistosomiasis Prevention and Control in China (2004—2015), a Bayesian interrupted time?series model was established to analyze the provincial reported incidence, time trend and seasonal variations of schistosomiasis in China at different stages. Results　The reported schistosomiasis cases were mainly concentrated in 5 provinces of Anhui, Jiangsu, Jiangxi, Hubei and Hunan and 2 provinces of Sichuan and Yunnan in China from 2004 to 2017, and the number of reported cases in endemic areas decreased gradually. The incidence of reported schistosomiasis cases predominantly peaked during the period from May to September in the marshland and lake regions, while no regular seasonality was seen in hilly regions. Bayesian interrupted time?series analysis showed the peak incidence of reported schistosomiasis cases in 4 provinces of Anhui, Hubei, Hunan and Jiangxi between May and September and in Jiangsu Province from July to November; however, no regular seasonal cycle was identified in hilly regions. The number of reported schistosomiasis cases showed a tendency towards an increase in 2 provinces of Hubei and Hunan from 2008 to 2014, with a minor peak during the period between March and April, and since 2015, the seasonality was not remarkable any longer in 3 provinces of Anhui, Jiangsu and Jiangxi with a decline in the incidence of reported schistosomiasis cases, while the seasonality remained in Hubei Province. Conclusions　The spatial?temporal characteristics of schistosomiasis in China, notably seasonality, vary at different control stages. Bayesian interrupted time?series model is effective to identify the spatial?temporal changes of schistosomiasis, and the schistosomiasis control strategy may be adjusted according to the spatial?temporal changes to improve the schistosomiasis control efficiency.

Key words: Schistosomiasis, Bayesian interrupted time?series model, Spatial?temporal distribution

摘要： 目的　分析2004—2017年中国血吸虫病报告病例时空变化规律，为进入不同防控阶段后的血吸虫病防治策略调整提供参考。方法　从公共卫生科学数据中心收集2004—2017年中国省级月度血吸虫病报告病例数据，描述血吸虫病报告病例时空分布；根据《全国预防控制血吸虫病中长期规划纲要（2004—2015年）》提出的阶段性目标建立贝叶斯中断时间序列模型，分析在不同防控阶段血吸虫病省级报告发病水平、时间趋势和季节性变化。结果　2004—2017年，我国血吸虫病报告病例主要集中在湖沼型流行区的安徽、江苏、江西、湖北、湖南等5省和山丘型流行区的四川、云南2省，且流行区血吸虫病报告病例数逐渐下降；湖沼型流行区血吸虫病报告病例发病高峰多在5—9月，山丘型流行区无规则季节性周期。贝叶斯中断时间序列模型分析显示，湖沼型流行区的安徽、湖北、湖南、江西4省血吸虫病报告病例发病高峰在5—9月，江苏省发病高峰在7—11月；山丘型流行区无规则季节性周期。2008—2014年湖北、湖南省血吸虫病报告病例数呈上升趋势，并出现了3—4月的小波峰；2015年后，随着血吸虫病报告病例数下降，安徽、江苏、江西3省发病季节性不再明显，而湖北省仍有一定季节性周期。结论　我国进入不同血吸虫病防控阶段后，血吸虫病报告病例时空分布特征尤其是季节性周期存在差异。贝叶斯中断时间序列模型可用于探究我国不同防控阶段血吸虫病报告病例时空变化规律，为针对其变化规律的防控措施调整提供参考，从而提高防控效率。

关键词: 血吸虫病, 贝叶斯中断时间序列模型, 时空分布